How to Measure Customer Experience in Financial Services

It’s time to listen closely to customers, build better models and look beyond your industry for comparisons.

As customer experience (CX) becomes a central battleground for financial services companies, a number of new questions have been hounding experience leaders, product owners, marketers and operations heads:

  • How do I measure what truly matters across the experience?
  • How might I align a mix of functions, business units and regions behind a unified view of what matters?
  • How might I motivate these groups to coordinate in delivering a superior experience where it matters most?

We have worked with clients across a broad spectrum of measurement sophistication. On the one end, some have spent millions on sophisticated measurement software, only to then struggle with translating their firehose of data into actionable insights. And on the other end, some still rely on a mix of CSV files and manually-generated reports across disparate systems – and struggle with finding meaning across the disjointed, hard-to-compare data.

“In our work, we have sought to make measurement more actionable by defining a unified CX measurement framework.”

We have found that there are five key tenets that can help companies measure CX in ways that provide clarity, improve decision-making, and ultimately drive business impact.

1. Start With What Matters Most To Customers

Leaders at large organizations will know all too well that it’s tempting to only measure interactions and transactions that sit within their domain. Yet, this common mindset produces an incomplete view of what truly matters to customers across their entire journey.

For a large financial institution in North America, we discerned what was meaningful to measure by starting with a customer-led view of what truly mattered to them across their end-to-end experience journey. We used qualitative techniques such as in-depth interviews and ethnographies to reveal pivotal moments across the experience. We then used quantitative research to sharpen our understanding of customer behavior at key moments and clarify how these influenced specific business outcomes.

2. Define a Unified Framework Across Levels and Functions

Most large organizations have multiple CX measurement frameworks, techniques, KPIs, and reporting mechanisms. While each of these might serve the purpose of distinct management levels and functions, they also create multiple and different versions of ‘what truly matters.’ This makes it particularly difficult for cross-functional teams to translate insights into action.

In our work, we have sought to make measurement more actionable by defining a unified CX measurement framework. Such a framework can typically span different management levels and functions while also identifying relationships across key measures that allow a more cohesive view.

With such a framework in place, senior executives, managers and front-line operators can all form a shared narrative about the firm’s CX performance, issues and opportunities. Executives can use high-level KPIs to measure the overall company CX priorities. Managers can use more detailed KPIs to define actionable milestones in service of the overall priority and allocate investments. Front-line operators can leverage a highly detailed subset of metrics to mobilize plans, establish service-level targets and track progress.

3. Build a Better Model with Leading and Lagging Indicators

The process of developing a unified CX measurement framework requires a sharp eye toward identifying the right measures that accurately describe customer impact and eventually business impact. Getting this part right often falls on ensuring we consider a broad range of data (ideally, data related to operational measures, customer sentiment/perception, customer behavioral response, and business outcome) as well as robust econometric models and analytics that connect CX measurement explicitly to financial value.

For example, in developing a model that derived relationships across different CX metrics for a large U.S. financial services firm, our data and analytics team made sure to:

  • Account for time dynamics where observations in one time-period are linked to observations in different time-periods
  • Capture interaction and endogeneity by allowing variables that are jointly determined to ensure estimates account for simultaneity and interaction of variables
  • Measure non-linear relationships and account for diminishing returns to ensure true influence is isolated
  • Control and capture influence of macro-economic changes and shocks that influence the business (e.g., shifts in interest rates or regulatory changes
  • Account for uncertainty based on probability — identifying expected outcome, what’s possible, and likelihood through Monte Carlo simulations

Our analytics team was also able to parse out what leading indicators managers should frequently look at (such as engagement, digital activity) and how these eventually predicted lagging indicators (such as customer acquisition, retention, advocacy) and ultimately financial performance. Most importantly perhaps, the model was translated into a what-if simulator that allowed our client to assess the likely financial impact from a variety of potential CX improvements.

4. Look Within, and Beyond, Your Industry for Comparisons

Competitive benchmarks are useful when trying to understand the areas of the experience to invest in. However, we believe it’s a mistake these days to compare your CX to just your competitors alone. Your customers are certainly going well beyond and comparing it with leaders across multiple categories – and this is especially true in sectors where satisfaction is systemically low.

For example, for a large global insurance provider, our research revealed that their CX scores within a key market in Asia were higher than most of their competitors – especially in parts of the journey that mattered most. However, a closer look also revealed that industry-wide scores in this market were significantly lower than other comparable markets, reflecting a more systemic, sector-wide level of customer dissatisfaction.

Despite temptations of proclaiming that they were providing a “leading experience,” managers at this insurer quickly agreed that they had no appetite for being “the best of the worst.” Instead, they recognized this as a clear opportunity to leap-frog their competitors and newer disruptors by doubling down on their relative strength in CX.

5. Invest Disproportionately in Defining and Developing a Measurement Governance Model

The most sophisticated of measurement strategies can end up failing if they are not accompanied by a governance model to deploy, maintain and ultimately act upon insights that evolve or transform the CX.

In our experience, a successful governance model typically solves for three key questions:

  • What people across what organization/functions will deploy, maintain and act on experience measurement reporting and insights?
  • What management processes will be required to drive systematic deployment, maintenance and actionable CX improvements?
  • What data, technology and interactive tools will be required to acquire, store and provision KPIs at the various levels of fidelity required across the organization


Ultimately, we believe that in the battle for winning on experience, firms that are able to combine a cogent experience measurement strategy with a robust governance model have a significant leg up. Such firms can see their end-to-end experience through the eyes of their customers. They can spot customer needs and opportunities in areas that matter most.

They can empower the right teams and executives with this insight quickly so they can act in real-time. And when they act, they can use customer insight to go beyond fixing what’s broken and deliver experiences that surprise and delight customers – and may even self-disrupt their model with more transformative innovation.

Learn more about how Prophet can help with customer experience to increase impact on your business.

Your network connection is offline.